Electroencephalographic response to music: Characterization using involvement indexes

Electroencephalogram (EEG) can be used to assess music-induced emotional responses evoked in listeners. To date, few EEG-derived indexes of involvement have been examined in this field. Thus, this study aims to investigate the use of EEG-derived involvement indexes as objective evidence of music-listening emotional effects, also comparing them with emotional states not affected by music. To do so, we have characterized and compared EEG-derived involvement indexes of a population under two different conditions: at rest without any auditory stimulus, and while listening to music. EEG data was acquired from 10 healthy subjects, as part of a freely available dataset on OpenNeuro, named “An EEG dataset recorded during affective music listening”. EEG recordings were processed using EEGLAB software. Preprocessing involved 0.5-70 Hz band-pass filtering, average re-referencing, and artifact removal via independent component analysis. Then, EEG rhythms were extracted, and 37 involvement indexes were derived as ratios of the spectral powers of two or more EEG rhythms. The Wilcoxon rank-sum test was applied to identify the most significant indexes for distinguishing between conditions. Results showed that 5 involvement indexes and 3 EEG channels, specifically O2, F8, and O1, were statistically significant in distinguishing between music listening and resting state. In addition, most of the indexes, if computed on specific scalp regions, were found to distinguish even among the different induced emotions (previously assessed by a Likert questionnaire). Our results, although preliminary, highlight the potential of EEG-based involvement indexes in distinguishing emotional conditions while listening to music and without music. This pilot study contributes to the field of affective computing and music therapy, supporting the development of EEG-based tools for emotion recognition and therapeutic interventions.